Amyotrophic lateral sclerosis is the most frequently occurring neuromuscular degenerative disorders, and has an obscure aetiology. Whilst major progress has been made, the majority of the genetic variation involved in ALS is, as yet, undefined. In this thesis, multiple genetic studies have been conducted to advance our understanding of the genetic architecture of the disease. In the light of the paucity of comprehensive genetic studies performed in Chinese, the presented study focused on advancing our current understanding in genetics of ALS in the Han Chinese population. To identify genetic variants altering risk of ALS, a genome-wide association study (GWAS) was performed. The study included 1,324 Chinese ALS cases and 3,115 controls. After quality control, a number of analyses were performed in a cleaned dataset of 1,243 cases and 2,854 controls that included: a genome-wide association analysis to identify SNPs associated with ALS; a genomic restricted maximum likelihood (GREML) analysis to estimate the proportion of the phenotypic variance in ALS liability due to common SNPs; and a gene- based analysis to identify genes associated with ALS. There were no genome-wide significant SNPs or genes associated with ALS. However, it was estimated that 17% (SE: 0.05; P=6×10-5) of the phenotypic variance in ALS liability was due to common SNPs. The top associated SNP was within GNAS (rs4812037; p =7×10-7). GNAS was also the most associated gene from the gene-based study (p =2×10-5). Based on GWAS data, a fragment- length and repeat-primed PCR was performed to determine GGGGCC copy number and expansion within the C9orf72 gene in the cohorts (1,092 sporadic ALS and 1,062 controls) from China.
A haplotype analysis of 23 SNPs within and surrounding the C9orf72 gene was performed. The C9orf72 hexanucleotide (GGGGCC)n repeat expansion (HRE) was found in three sALS patients (0.3%) but not in control subjects (p = 0.25, Fisher’s exact test). Two cases with HRE did not harbor four risk alleles that have previously been determined to be strongly associated with ALS in Caucasian populations. The presence of risk alleles (including rs2814707 and rs384992) of the 20-SNP consensus risk founder haplotype in Caucasians demonstrated that two of the three cases shared a novel haplotype carrying HRE. The low frequency (1.8%) of the 20-SNP consensus risk haplotype and the distinct allele distribution in Chinese ALS patients compared to Caucasian populations indicates that the C9orf72 HRE is not from the same single founder haplotype involved in Caucasian populations. In addition, using next generation sequencing, a population-specific mutational spectrum of ALS demonstrated that mutations in SOD1 are the most common causative mutations in fALS in the Han Chinese population.
To identify further genetic associations, a novel strategy based on functional annotations and current understanding of ALS for specifically filtering candidate genes has been suggested. These findings broaden the spectrum of causative mutations in ALS and are essential for optimal design of strategies of mutational analysis and genetic counselling of Han Chinese ALS patients.
To conclude, the presented study with the largest non-Caucasian ALS cohorts showed the heterogeneity of ALS in the Chinese Han population and laid the foundation for further genetic studies.